doc: gp_pred
Posterior mean and variance of GP given the kernel matrices and the Bayesian inferance
Syntax
[mu, sigma2] = gp_pred(Kts, dKss, BayesInv) [mu, sigma2] = gp_pred(Kts, dKss, BayesInv, Ht, Hs)
Arguments
- Kts matrix (nt, ns) of kernel between the points of Xt and Xs
- dKss matrix (ns, 1) of diagonal kernel between the points of Xs
- BayesInv structure array returned by gp_inf(Ht, Ktt, Yt, noise)
- Ht matrix (nt, b) of basis data for Xt
- Hs matrix (ns, b) of basis data for Xs
Outputs
- mu matrix (ns, 1) of posterior mean
![$E[f(X_s) \mid X_t, Y_t]$](gp_pred_eq75965.png)
- sigma2 matrix (ns, 1) of posterior variance
![$V[f(X_s) \mid X_t, Y_t]$](gp_pred_eq20100.png)